|
||||||||||
|
DOI: 10.17587/it.25.490-501 A. V. Vishnekov, Ph. D., Professor, e-mail: avishnekov@hse.ru, E. M. Ivanova, Ph. D., Associate Professor, e-mail: emivanova@hse.ru, K. E. Basova, master's student, "Computer systems and networks", e-mail: kezubakina@edu.hse.ru, E. O. Vetelina, master's student, "Computer systems and networks", e-mail: eovetelina@edu.hse.ru, National Research University Higher School of Economics, Moscow, 123458, Russian Federation This article aims at the improvement of the engineer training quality in the field of "computer Science and engineering" using on-line and blended learning technology. We consider studying of modern computers architecture as an example. The complexity of the modeling object, the multi-faceted process of studying the computer architecture, the limited period — are the factors that determine the difficulty of the problem. A computer simulator allows you to explore the stages and features of the calculation process in dynamics with the necessary detail. It can't be done any other way. Also, computer modeling creates a risk-free environment that allows you to safely and repeatedly apply and analyze possible scenarios, decomposition of computer nodes and configuring its components for different modes of operation. This is either impossible to perform on real devices or may result in computer failure under students' inept actions. The paper discusses the problems of the computer units and devices simulating, the requirements for models in terms of solving educational and research problems. We propose a structure of the simulating complex, visualization methods of computer architecture components and principles of computer system device's operation, algorithms descriptions of the modeling programs. Since the computer architecture implies a whole complex of concepts, including functional and structural organization, software and hardware interaction algorithms, the considered simulating complex consists of several simulators: step-by-step operation of the superscalar processor core; cache coherence protocols in multiprocessor (multi-core) systems; algorithms for fast address translation with the help of Translation Lookaside Buffer (TLB); various technologies for multithreaded data processing; pipeline processing of the machine instructions stream; memory paging algorithms; cache mapping algorithms; cache replacement algorithms; etc. The main functions and features of each model are considered. An important advantage of the proposed simulating complex is the ability to vary the parameters of models of specific computer devices, which makes it possible to explore the resource/time characteristics of data processing by computer system devices. P. 490–501 |